Valve sizing for flashing liquids 4

[Renoyd]

Hello,

What would be the proper methodology for sizing a control valve with liquids flashing downstream?

Fisher recommends allowable pressure drop method that results in a larger valve than using actual pressure drop. The question is, how can I get the required pressure drop if the much lower allowable pressure drop is used for sizing?

Any idea or commends would be appreciated.

 

[rmw]

Fisher has extensive valve sizing help on their website including provisions for flashing service. If memory serves, it is their Catalog 12. If not, it will be in the same group of technical data where Cat 12 is to be found. It may be the next one or the third one listed. I think however, the third one is all about noise.

Sorry, while I refer to them often, I downloaded them long ago (onto my work computer), so I don't have to go to where I am sending you so I don't know the path right off hand (and I will leave looking it up to you).

rmw

 

[hsbcn]

Asco also has some useful tools on their site:

http://www.ascovalve.com/Applications/ValveSizing/ValveSizing.aspx

Hope this helps!

My focus: Alloy Valves Duplex Valves Super Duplex Valves Monel Valves Inconel Valves [link titanium-valves.com]Titanium Valves[/url]

http://www.6-moly-valves.com

 

[JimCasey]

>>how can I get the required pressure drop if the much lower allowable pressure drop is used for sizing?<<

When flashing occurs in a valve, the flow is choked. "Choked" is an unfortunate term but when choking happens the actual downstream pressure is independent of the flowrate. The effective DP is P1-the pressure at which the flow becomes choked.

Download the sizing program from a major control valve manufacturer. Flowserve (Valtek) , Emerson (Fisher), Dresser (Masoneilan) all offer free software.

Water expands around 1000 times its liquid volum when it flashes completely, although typical flashing usually vaporizes around 15% of the mass of the liquid. That's enough to make the downstream flow very fast. Rotary valves serve well for flashing because the outlet is straight out and the high-speed droplets don't scour the valve internals. It's a good idea to calculate the velocity downstream of the valve and to transition to a larger pipe diameter to keep the velocity within reasonable limits so the downstream system erosion is managed. Also, Chromium-bearing alloys resist flashing erosion a lot better than carbon steel, so select piping accordingly.

Calculate the percent flash by setting up an isenthalpic balance:

Hf1=XHg2+(1-x)HF2 Where H is Enthalpy, F is Liquid, G is vapor, 1 is upstream and 2 is downstream. X is the percent of vapor phase.

Once you have the percent flash you can calculate the specific volume, then the velocity.

 

[Renoyd]

Jimcasey,

When you say effective dp, is that allowable dp used for sizing?

If P1=150 psig, p2 needs to be 10 psig (process requirements). However, the allowable dp is only 20 psi. Then 20 psi is used to size the valve, say, passing the total flow at 70% opening. So, the valve will be 70% open with 20 psi dp. How can I get P2 to 10 Psig?

 

[rmw]

What he said was that your downstream pressure will be the pressure at which the flow becomes choked. In other words, if the flow becomes choked at... uhmmmm, let's say 90 psi, then 90 psi is your downstream pressure.

In a situation like this, you would have to have multiple let down valves, or choose a valve that has a trim capable of taking a large pressure drop without becoming choked. Those valves usually work on the basis of multiple let down stages within the trim itself so that no stage becomes choked.

Fisher has valves that will do it, but in my experience, there are others that specialize in this type of service and do a better job of it. (Not to knock Fisher, I bought 3 of their valves today and there were 4 bidders.)

rmw

 

[PaoloPemi]

with mixtures you can utilize tools as Prode Valves which solve an adiabatic flash (or use a process simulator), with the common pure fluids the Masoneilan method gives reasonably results but you need parameters as FL etc. which are not easy to find, I prefer to ask the manufacturer to do all the calc's and have the responsability for the results, the same for the max noise.

 

[Scotsinst]

Renoyd, your question is incorrect. The downstream pressure is 10psig. The valve flashing calculation does not affect the downstream pressure.
Imagine you were discharging to atmosphere. Downstream pressure is 0 and will always be 0. Like JimCasey mentioned; the effective pressure drop (choked) is used for calculation of flowrate, not to calculate downstream pressure.
rmw, I'm afraid your post is incorrect. Multiple letdown valves or multistep trim is only required for high-pressure (>500psip)cavitation elimination or cavitation/flashing with multi-component streams. You can only eliminate flashing by reducing the vapour pressure (cooling) or increasing the downstream pressure to above vapour pressure.
For small Cv's select a globe valve with reduced stellited trim (to allow expansion within the valve) or for larger Cv's Camflex style rotary valves installed flow-to-close are good and swage up your downstream line to minimize erosion.
PaoloPemi; I'm afraid your answer scares me because you are relying on someone who may not be competent. In the good old days Fisher and Masoneilan had good reps who provided great support but they have cut resources just like everyone else. I recommend you study and try hard to understand exactly what is being proposed by vendors.

 

[PaoloPemi]

Scotsinst, I see the point but unless you are very competent you can't do any reasonably good two phases calc without specific valve parameters which are (usually) available only at manufacturer's tech dept... so I think it's better to ask them and verify...

 

[rmw]

I stand by my post.

rmw

 

[Scotsinst]

PaoloPemi, I agree that you can get different vendors to size 2-phase flow for a sizing check but make sure your process engineer does a flash calc to check the vapor/liquid fractions. Many times the vendors overestimate valve Cv due to flashing. The effects of subcooling also reduce the required size. My point is that you need to be wary of relying on vendors and what kind of 'guarantee' are they going to give you? Valve vendors are generally not process engineers.

rmw, don't take it personal but if someone suggested using multiple valves for 150psi pressure drop they would not be taken seriously. In addition, your solution of multiple valves will not prevent flashing.

 

[rmw]

Scot... Nothing personal about it. Go back and reread the post and the two that preceded it. I don't think I used the word flashing at all in my post. I was addressing the issue of choked flow answering a question the OP asked the previous poster about his words on the topic. I picked the value of 90 out of the air as an example only. I have no idea what the upstream pressure might have been, just the choked flow downstream pressure. I was illustrating a principle, not designing his system.

I still stand by my post and will continue to do so, although I will not continue this diatribe.

rmw

 

[JimCasey]

>> Valve vendors are generally not process engineers. <<

True enough, but Valve application engineers are, in my experience, expert at what they do and a valuable resource to process engineers.

I wish I knew how many times process engineers had sent me data sheets where they neglected to specify or omitted the downstream pressure (What size valve do I need for 100 psi steam?) , or specified the same upstream pressure and downstream pressure across a wide flow range.

 

[Scotsinst]

rmw: I apologize if my post was offensive to you. I thought you were answering the posed question.
JimCasey: I used to have your faith in valve application engineers (I was a valve application engineer for Masoneilan and Fisher many years ago) and I'm sure if you work in Houston, London or other large engineering centers they still have good application engineers but it's sadly not the case everywhere. Good luck with dealing with your process engineers.

 

[Renoyd]

I thank everyone for your valuable input.

What exactly is the definition of flashing here? Think about the following two cases:

1) Dehy contactor level control valve: some dissolved gas flashes out from the glycol downstream of the valve.

2) Refrigeration chiller level control valve: propane liquid is let down to almost atmosphere presure and C3 vapor flashes out from the C3 liquid.

The second case is definately a flashing case for valve sizing. How about the first one?

 

[PaoloPemi]

in both cases at outlet there will be some gas by flashing (the process of liquid vaporizing into gas), in the FIRST case (Glycol + Natural gas) the exact behaviour is more difficult to evaluate as we have a mixture, you can model this as an adiabatic flash (see my previous post for suitable software tools) or simply as isothermal flash, then calculate the amount of gas and liquid flows and have some averaged values to use in control valve formulation. With mixtures if gas content and dp are high you should also verify the condition of choking flow.